Smart robotics for high added value footwear industry Deliverable

Transcription

1 Project ref. no Project acronym Project full title Type Dissemination level Contractual date of delivery Actual Date of Delivery ROBOFOOT Smart robotics for high added value footwear industry Deliverable CO M20 M26 Title of document D6.1 Definition of the Demonstrator Version V2.0 Number of pages Partner Responsible Other Contributors Author INESCOP TEKNIKER, DFKI, CNR-ITIA, AYCN INESCOP Keywords: Abstract Demonstrator, Robot, inking, polishing, last removal, programming. Description of the DEMONSTRATORS that have been designed, in particular that will be presented in the SIMAC trade fair in October 2012

2 Document History Ver. Date Changes Author Structure of deliverable INESCOP Revised structure of deliverable TEKNIKER Added contributions INESCOP INESCOP and AyCN contributions INESCOP+AyCN Contribution from CNR-ITIA TEKNIKER Integrated version with additional contribution from CNR-ITIA CNR-ITIA Version ready for last revision before submission TEKNIKER Version submitted to the EU TEKNIKER New version submitted to the EU, including the description of the BIEMH12 demonstrator and the permanent demonstrator at partners facilities TEKNIKER ROBOFOOT_D6.1-v2.0 l Page 2 of 21

3 Content 1 List of figures and tables Figures Tables Introduction SIMAC Demonstrator description Stand description Cell description Sequence of operations Other elements presented Programming environment Off-line programming: BasicCAM ROBOmove postprocesor D Scanner AILA Supporting material BIEMH12 demonstrator Permanent demonstrators...21 ROBOFOOT_D6.1-v2.0 l Page 3 of 21

4 1 List of figures and tables 1.1 Figures Fig. 1 General overview of the robotic cell (I)... 6 Fig. 2 General overview of the robotic cell (II)... 6 Fig. 3 Layout of cell, dimensions... 7 Fig. 4 Inking cabinet for the Demonstrator... 8 Fig. 5 Ink drying buffer... 8 Fig. 6 Polishing machine... 9 Fig. 7 Last opening station...10 Fig. 8 Sequence of operations...10 Fig. 9: General scheme for trajectory generation/control...11 Fig. 10 Path definition for inking...12 Fig. 11 Path definition for polishing...12 Fig. 12 ROBOmove simulation (I)...13 Fig. 13 ROBOmove simulation (II)...13 Fig. 14 Laser scanner 3D single module model with working area...15 Fig. 15 ITIA-CNR Laser scanner system 3D...15 Fig. 16 ITIA-CNR Laser scanner system 3D with working volume...16 Fig. 17 laser scanner 3D single column with working volume...16 Fig. 18 point clouds on cad reference system (black) and on scanner reference system (pink) before and after registration procedure...17 Fig. 18 TEKNIKER booth (right of the stand shared with Ideko)...19 Fig. 20 Details of the ROBOFOOT Demonstrator in the stand: the robot, manovia,...20 Fig. 18: Demonstrator at TEKNIKER...21 Fig. 19: Demonstrator at INESCOP...21 Fig. 19: Roughing demonstrator at CNR-ITIA Tables No se encuentran elementos de tabla de ilustraciones. ROBOFOOT_D6.1-v2.0 l Page 4 of 21

5 2 Introduction The aim of this document is to define the demonstrator that will be used for public dissemination. Three mechanisms have been developed: Robotic cells permanently available at INESCOP, CNR-ITIA and TEKNIKER An adapted version to be presented in BIEMH12 An adapted version to be presented in SIMAC12 The main effort corresponds to the SIMAC Trade Fair, the most important fair in the footwear sector. It takes place every two years in Bologna. ROBOFOOT will present a demonstrator of some of the results achieved during the first months of the project in that fair. This document describes in detail the elements that are part of the SIMAC demonstrator: A robotic cell with inking, polishing and last opening operations Path programming environment: o Basic CAM o Off-line path adjustment Bimanual, multifinger robotic packaging ROBOFOOT_D6.1-v2.0 l Page 5 of 21

6 3 SIMAC Demonstrator description The key element of the DEMONSTRATOR is the robotic cell, based on that implemented for PIKOLINOS, i.e. the cell including inking, polishing and last removal operations. 3.1 Stand description A stand has been allocated for the presentation of ROBOFOOT results. The main element of the stand is the robotic cell. The stand will be presented in a specific area devoted to robotic applications. The layout of the cell is presented in the following pictures: Fig. 1 General overview of the robotic cell (I) Fig. 2 General overview of the robotic cell (II) ROBOFOOT_D6.1-v2.0 l Page 6 of 21

7 Fig. 3 Layout of cell, dimensions 3.2 Cell description Inking cabinet A new cabinet has been designed for the Demonstrator to facilitate visitors having a good vision of the inking process. It includes a new air suction system. ROBOFOOT_D6.1-v2.0 l Page 7 of 21

8 Fig. 4 Inking cabinet for the Demonstrator Ink drying Buffer An 8 position buffer will be used to allow drying the ink before polishing. Fig. 5 Ink drying buffer ROBOFOOT_D6.1-v2.0 l Page 8 of 21

9 Polishing machine A special polishing machine has been designed and manufactured to be integrated in the cell. This new machine is designed without protections in order to allow seeing the full process during the exhibition. Fig. 6 Polishing machine Last opening station At the end of the polishing process the last will be opened in the workstation designed to this aim. As it will be presented in a fair, in order to be an unattended process that repeats cyclically, instead of leaving in a conveyor as would be the real process the robot will close the last and put it again in the buffer. ROBOFOOT_D6.1-v2.0 l Page 9 of 21

10 3.3 Sequence of operations Fig. 7 Last opening station As explained before, the sequence of operations will be cyclic for the eight shoes: Take the shoe from the buffer (8 positions) Inking Polishing Last opening (and close) Take it back to the buffer SHOES x 8 BUFFER 2 SHOES x 8 1 POLISHING INKING LAST REMOVAL Fig. 8 Sequence of operations ROBOFOOT_D6.1-v2.0 l Page 10 of 21

11 4 Other elements presented 4.1 Programming environment It will be presented the programming approach adopted by ROBOFOOT, described in D4.1, in particular the trajectory definition, the postprocessor and the OFF-line program adjustment using the 3D laser scanner. Fig. 9: General scheme for trajectory generation/control Off-line programming: BasicCAM BasicCAM is a software developed in the project to help users defining the path the robot has to follow to perform different operations as well as some parameters related with the operation. It will be presented how this software is used for inking and polishing operations. ROBOFOOT_D6.1-v2.0 l Page 11 of 21

12 Fig. 10 Path definition for inking ROBOmove postprocesor Fig. 11 Path definition for polishing The path generated with BasicCAM is translated in an executable program in the robot language, in our case in PDL2 format. This process is done by means of the ROBOmove postprocessor created by QDESIGN. ROBOmove takes care of: Robot kinematic and reachability: The translation of a point in the space into a point reachable by the robot tool means the solution of the inverse kinematic of the robot model, i.e., the mapping of that point in the space into a set of 6 joint angles (one for each robot axis). The solution to the inverse kinematic equations is provided by ROBOmove. If at least a solution exists ROBOFOOT_D6.1-v2.0 l Page 12 of 21

13 the point is reachable by the robot and ROBOmove informs the user about it, as well as if any point is outside the robot working area. Management of the robot redundancy: The relationships between joint space and Cartesian space coordinates of the object held by the robot are in general multiple-valued: the same pose can be reached by the serial arm in different ways, each with a different set of joint coordinates. In other words, the solution of the kinematic chain of the robot gives more than one solution for the 6 joint angle values the robot must have to reach a point in the space. ROBOmove solves the kinematic of the robot giving an unique solution. This is done managing the redundancy, acting on the wrist pose. Given a point and a direction for the tool, the robot can move the tool to that position/orientation in more than one pose: Controlling the pose of the wrist (the 4th, the 5th and the 6th angles), this redundancy can be managed. The solution suggested by ROBOmove to manage this 6 th DOF is the rotation around the axis of the tool. The following pictures show the example of a robot holding the last to be milled. The tool touches the same point (near the last-tip) in the 4 different pictures, but in each picture the posture of the robot is different and is controlled by a different value of rotation around the tool axis. Fig. 12 ROBOmove simulation (I) Management of singularities: Fig. 13 ROBOmove simulation (II) ROBOFOOT_D6.1-v2.0 l Page 13 of 21

14 Some robot poses cannot be reached in a unique way because there are infinite solutions of the kinematic chain in that point. If the robot path pass near one of this singular points the behavior of the movement can t be predictable around it. ROBOmove avoids moving the robot in the vicinity of those points. Simulation and Collision Detection: The offline simulation of the machining is important in order to detect possible collisions between the robot and/or end effector and/or other part in the environment. ROBOmove offers a 3D simulation of the plant scenario: robot, external axes, spindle, tool, workpiece, any fixture and any other significant element of the robotic cell. The user has the opportunity to introduce new graphics on the scene. Moreover, the ROBOmove simulation windows gives other important information: The machining cycle time. The displayed tool path is colored depending on if the points are reached by the robot or not, as well as if the points match the conditions of singularities of the robot. ROBOmove has got some Settings Tool and Utilities by which: The user can set: o o o The user can: o o Piece Positioning Frame Initial robot posture Velocity setting modify the points coordinates /normal set by BasicCAM, if needed add some points useful for the machining (i.e, transition points). o use appropriate strategies for reducing the number of points in the APT / ISO language program. Robot code Output Customization ROBOmove has been customized in order to managed special commands (not foreseen in the standard APT language) needed to manage particular operation. For instance, some non-standard special commands can be used to insert some robotlanguage-instructions directly in ROBOmove in operations as polishing and roughing D Scanner The LSS developed allows the instrument for last+gasper digitalization, both HW and SW, it has been fully and internally developed; the solution is a laser scanner able to work, in a reliable way, in an uncontrolled environment and characterized by an open and modular architecture. The solution is a linear scanner; the multiple head allows the whole acquisition of the subject under digitalization in only one linear movement; a single scan is enough for digitalizing the subject entirely. For the acquisition of the last coupled with a gripping device the requirements were: working volume equal to 300x300x600 mm3 accuracy equal or less than 0.1 mm ROBOFOOT_D6.1-v2.0 l Page 14 of 21

15 acquisition time equal or less than 30 sec. To reach the final configuration different conceptual designs of LSMs and LSS have been developed and different tests for the identification of suitable components (laser, optic, camera etc.) have been performed. The optimal solution selected consists of: LSMs: each LSM is provided by one camera (752 x 480 pixels) and a laser emitter (Class 3R) 1 linear axis that moves the target (last coupled with the gripping device) The target motion and the acquisition are synchronized and managed by the same PC embedded. Fig. 14 Laser scanner 3D single module model with working area Fig. 15 ITIA-CNR Laser scanner system 3D Final solution: 4 linear laser scanner heads with: Cam Pt Grey Fire-Fly MV Usb 752 each Optic Fujinon da 6mm DF6HA-1B each Filtri Midwest BP each Laser Globallaser Lyte-MV 660, 5mW each Interax Cam-laser 150mm 1 linear axis run 1000mm ROBOFOOT_D6.1-v2.0 l Page 15 of 21

16 Fig. 16 ITIA-CNR Laser scanner system 3D with working volume Fig. 17 laser scanner 3D single column with working volume Thus, measuring the single-shoe or the single-last in the production system seems, by now, the only solution in order to allow the introduction of robot inside the production line, paving the path for autonomous adaptation of the nominal robot part program to the actual geometrical characteristics of the shoe being manufactured The design followed is based on modular design, allowing the use of low-cost components, and directly customized on the specific application s requirements. The laser scanner system has solved the problem of the measure of the actual coupling between the last and the gripping device that is usually characterized by a poor accuracy. To perform the alignment between the CAD model and the acquired cloud of points a procedure, based on Iterative Closest Points (ICP) optimization methods, has been developed and directly integrated in the laser scanner device. As a result of this procedure a roto-translation matrix is communicated directly to the robot controller. ROBOFOOT_D6.1-v2.0 l Page 16 of 21

17 The two sets of data input are: the 3D points of the last coupled with the gripping device in the CAD model Xpc-cad; the 3D points acquired by the LSS Xpc-laser The goal of the alignment procedure, consists of identifying (registration process) the 3D rigid transformation (rotation Rpc-cad and translation Tpc-cad) between the CAD nominal model and the 3D object reconstructed by the LSS (each point is expressed in the same Cartesian frame). Iterative algorithms allow an easy and cost-efficient solution; target accuracy can be fixed by defining a user-selected tolerance threshold τ>0 that identify the exit condition for the algorithm (iteration terminates when the mean-square error is below the threshold). Fig. 18 point clouds on cad reference system (black) and on scanner reference system (pink) before and after registration procedure An automatic procedure that modifies off-line the robot part-program allows higher accuracy in the tool path planning so reducing the unwanted-constraint forces when the shoe is brought in contact with tools by a robot. For the evaluation of the efficacy of this method many test have been performed; the experiments consist on moving shoe in contact with different tool with and without the off-line correction of the path and performance evaluation is given by the measurement of the contact forces. Results and more details can be found in D AILA DFKI will provide project flyers for the Fair which explain the contribution of DFKI to the project ROBOFOOT in terms of non-rigid part manipulation by means of multifinger-dual-arm robots. Moreover, a video will be shown in which the contribution of DFKI to the project is demonstrated. In this case, the video will show the dual-arm robot AILA autonomously packing a pair of shoes into a box. The robot learned to perform this task after a few demonstrations given by a human operator. The approximately 3-minute video will consist of real sequences of the robot at DFKI Laboratories combined with screen captures of the simulations performed during the learning process. ROBOFOOT_D6.1-v2.0 l Page 17 of 21

18 4.3 Supporting material Different dissemination material will be ready for distribution and presentation in the fair -Poster and leaflet -Videos of partial results ROBOFOOT_D6.1-v2.0 l Page 18 of 21

19 5 BIEMH12 demonstrator An adapted version of the cell implemented at TEKNIKER s facilities has been prepared for its presentation at Machine Tool Fair BIEMH12 that takes placed every two years in Bilbao (Spain), considered the most important Spanish Industrial fair. The fair took place from May 28 th to June 2 nd TEKNIKER will present this cell in a booth. Fig. 19 TEKNIKER booth (right of the stand shared with Ideko) The demonstrator includes also a big screen where the process is presented to visitors: The visual servoing process: the images captured by the camera in the robot arm. The image treatment and the intermediate pose identification is shown. The inspection process: the image captured and the detected defects. The description of the sequence presented is as follows: The robot identifies the pose of the shoe in the manovia According to the shoe, different actions are done: o Inspection o Polishing operation Finally the shoe is placed back in the manovia The sequence is repeated for the different shoes. ROBOFOOT_D6.1-v2.0 l Page 19 of 21

20 Fig. 20 Details of the ROBOFOOT Demonstrator in the stand: the robot, manovia, ROBOFOOT_D6.1-v2.0 l Page 20 of 21

21 6 Permanent demonstrators According to the plan, three different demonstrators are permanently available at INESCOP, CNR-ITIA and TEKNIKER o INESCOP. It includes the manovia and inking, ink drying, polishing and last opening operations. Visitors can see the full process according to the following sequence: The robot takes the shoes from the manovia The grasping orientation is changed in the rotating station The robot performs the inking operation in the cabinet The shoes are placed in the buffer for inking drying Once dried, the robot takes the shoes and performs the polishing operation Finally the robot opens the last o TEKNIKER. It includes the manovia and polishing and inspection operations. In the coming months the packaging prototype will be included as well. Visitors can see the full process according to the following sequence: The robot takes the shoes from the manovia The grasping orientation is changed in the rotating station Two type of shoes are placed in the manovia and according to them the following operations are performed by the robot: o The robot performs the polishing operation o The robot inspect shoes quality (nails, adhesive, roughing, texture). o CNR-ITIA. The roughing operation is integrated in the cell. It includes a COMAU robot and the roughing station. Fig. 21: Demonstrator at TEKNIKER Fig. 22: Demonstrator at INESCOP Fig. 23: Roughing demonstrator at CNR-ITIA ROBOFOOT_D6.1-v2.0 l Page 21 of 21